Skip to Main Content

AI Tools for Academic Literature Research

This guide includes examples of available AI-powered tools to assist with scholarly literature searches and reference management plus ethical considerations and warnings.

Warning!

Important Terms of Use Restrictions

  • Several subscription-based academic literature databases (i.e. those provided by the Library such as EBSCO products) restrict the use of database content with external AI tools.
    • EBSCO's (2024) License Agreement states, "Licensee and Authorized Users may not use artificial intelligence tools or machine learning technologies with any of the content included in the Databases or Services for any purpose."

Ribbon sign blocking entry to an area. Sign reads DANGER.

 

 

 

 

 


View larger image

  • Elsevier includes the following copyright statement on its site pages: "All rights are reserved, including those for text and data mining, AI training, and similar technologies. For all open access content, the relevant licensing terms apply."

Finding Open Access Materials

Workarounds to Avoid Database Terms of Use Restrictions

  • Look for an openly available copy of the article/chapter, etc., outside the proprietary database
    • Do a web search (Google, Google Scholar) using the article title to search for pre-published copies
    • Search preprint servers such as ArXiv and others. (ASAPbio offers a robust spreadsheet of information about different preprint servers.)
  • When using an AI-enabled literature review support tool (e.g. Elicit), ask a research question or link (using the doi or permalink) to an article of interest rather than uploading a document file.

NotebookLM (Google)

NotebookLM Overview

  • NotebookLM allows users to curate collections of resources (text, images, audio, video, markdown), up to 50 per collection, and then ask questions or prompt in other ways, to analyze the information in that collection.
  • The free version allows a user to create up to 100 notebooks, each containing as many as 50 sources. The size of each source is limited to 500,000 words.
  • Automated prompts are available at the click of a button. These include options to create a Study Guide, FAQ, Briefing Doc, FAQ, Timeline, and Audio Overview.
  • According to the website, "NotebookLM does not use your personal data, including your source uploads, queries, and the responses from the model for training."
  • NotebookLM Plus is the paid version that offers 5x the number of audio overviews, notebooks, queries, and sources per Notebook. It also provides shared Notebook access for teams, usage analytics, and "additional privacy and security."
  • When the user inputs their own query, NotebookLM generates a response with in-text references to the source materials. The user can save NotebookLM's responses as Notes to their dashboard.
  • The Studio feature allows users to create audio overviews in the style of a podcast with two hosts. A customization option provides for specialized instructions, such as intended audience, sources to emphasize, etc.

 

 

 

 

Adding Sources to NotebookLM

  • Users on the free plan can add up to 50 sources per Notebook.
  • Accepted file types include PDF, .txt, Markdown, and .mp3. Additionally, Google docs, Google Slides, links to websites and videos, and pasted text can also be used.

Shows NotebookLM page with option to upload sources of different types.

View larger image

  • After uploading sources, the Notebook dashboard becomes the user interface for interacting with the materials.

Notebook dashboard showing source file list in left panel, summary and overview of source content plus the open prompt box, plus options for generating and audio file and a briefling doc. in the center panel,. The right panel shows the Studio which allows for podcast generation including customization features. Also in the right panel, are auto-generate options for a Study guide, FAQ, and Timeline. Finally, the user's previous queries and responses are saved as notes.

View larger image

  • Notebook dashboard showing the three-panel view.
  • The left panel shows the list of sources added with options to select them.
  • The center panel provides a summary plus the open prompt box, plus options for generating and audio file and a briefling doc. in the center panel,. The right panel shows the Studio which allows for podcast generation including customization features. Also in the right panel, are auto-generate options for a Study guide, FAQ, and Timeline. Finally, the user's previous queries and responses are saved as notes.

 

Using NotebookLM to Create a PodCast

After you have uploaded source documents into a Google Notebook, you can use the Studio feature to generate a podcast. The podcast enables two AI hosts to discuss the material that you uploaded as if they were having a conversation about it. It is presented in the style of a podcast with the hosts engaging in a natural language, back-and-forth exchange. 

Shows GoogleNotebook LM's Studio for Creating Podcasts. A pink oval encircles the word Studio and a gold arrow points to the Interactive Mode (Beta) button

View larger image

The image indicates that a 21-minute podcast was generated called AI Trends, Applications, and Ethical Considerations. The Interactive Mode (BETA) seems to allow users to connect to the podcast and ask questions (by speaking). This would then prompt the hosts to respond to specific questions. This feature may not be widely available.